Polar Biology

, Volume 32, Issue 7, pp 1055–1065

Biogeochemical conditions and ice algal photosynthetic parameters in Weddell Sea ice during early spring


    • Antarctic Climate and Ecosystems Cooperative Research Centre
  • S. Papadimitriou
    • School of Ocean Sciences, College of Natural SciencesBangor University
  • D. N. Thomas
    • School of Ocean Sciences, College of Natural SciencesBangor University
  • L. Norman
    • School of Ocean Sciences, College of Natural SciencesBangor University
  • G. S. Dieckmann
    • Alfred Wegener Institute for Polar and Marine Research
Original Paper

DOI: 10.1007/s00300-009-0605-6

Cite this article as:
Meiners, K.M., Papadimitriou, S., Thomas, D.N. et al. Polar Biol (2009) 32: 1055. doi:10.1007/s00300-009-0605-6


Physical, biogeochemical and photosynthetic parameters were measured in sea ice brine and ice core bottom samples in the north-western Weddell Sea during early spring 2006. Sea ice brines collected from sackholes were characterised by cold temperatures (range −7.4 to −3.8°C), high salinities (range 61.4–118.0), and partly elevated dissolved oxygen concentrations (range 159–413 μmol kg−1) when compared to surface seawater. Nitrate (range 0.5–76.3 μmol kg−1), dissolved inorganic phosphate (range 0.2–7.0 μmol kg−1) and silicic acid (range 74–285 μmol kg−1) concentrations in sea ice brines were depleted when compared to surface seawater. In contrast, NH4+ (range 0.3–23.0 μmol kg−1) and dissolved organic carbon (range 140–707 μmol kg−1) were enriched in the sea ice brines. Ice core bottom samples exhibited moderate temperatures and brine salinities, but high algal biomass (4.9–435.5 μg Chl a l−1 brine) and silicic acid depletion. Pulse amplitude modulated fluorometry was used for the determination of the photosynthetic parameters Fv/Fm, α, rETRmax and Ek. The maximum quantum yield of photosystem II, Fv/Fm, ranged from 0.101 to 0.500 (average 0.284 ± 0.132) and 0.235 to 0.595 (average 0.368 ± 0.127) in the sea ice internal and bottom communities, respectively. The fluorometric measurements indicated medium ice algal photosynthetic activity both in the internal and bottom communities of the sea ice. An observed lack of correlation between biogeochemical and photosynthetic parameters was most likely due to temporally and spatially decoupled physical and biological processes in the sea ice brine channel system, and was also influenced by the temporal and spatial resolution of applied sampling techniques.


Sea iceAntarcticWeddell SeaBiogeochemistryPAMPhotosynthesisIce algae


The annual advance and retreat of sea ice around the Antarctic continent is a key factor of Antarctic marine ecosystem function and plays a pivotal role in the biogeochemical cycles of the Southern Ocean (Brierley and Thomas 2002). The sea ice cover greatly affects the exchange of energy and mass between the atmosphere and the sea. It also strongly affects pelagic production due to its effect on the under-ice light regime and coupled physical-biological processes at retreating ice edges during spring, including water column stratification, release of nutrients and seeding of the water column with ice algae (Lizotte 2001; Thomas and Dieckmann 2003).

Sea ice provides a vast low-temperature habitat for ice algae and is characterised by steep vertical and horizontal gradients in temperature, salinity, light and nutrients (Eicken 1992). Ice-associated primary production can contribute up to 25% to the overall production of ice-covered waters in the Southern Ocean (Arrigo and Thomas 2004; Lizotte 2001). The ice algae colonise distinct habitats at the surface, interior and bottom of sea ice floes (Arrigo 2003; Horner et al. 1992). The bottom communities of coastal fast ice have been intensely studied and often show high algal biomass due to their contact with the under-ice water allowing nutrient replenishment that may support algal growth. While the distribution of internal and surface sea ice communities varies around the continent (Grose and McMinn 2003), internal communities are generally found in offshore pack ice and are especially important in the Weddell Sea (Arrigo 2003; Arrigo et al. 1997). The interior of sea ice, consisting of two phases brine (hypersaline water inclusions) and solid ice, resembles a semi-enclosed habitat, which, depending on the physical properties of the ice, can be isolated from the underlying water column. A threshold of ≤5% brine volume fraction is considered to inhibit seawater percolation and transport of nutrients into the sea ice interior (Golden et al. 1998). Subsequent nutrient demand of ice algae in the ice interior may exceed resupply and, in case of insufficient heterotrophic nutrient remineralisation, will lead to depletion of the major nutrients nitrate, silicic acid and phosphate. The photosynthetic activity can also drive dissolved inorganic carbon concentration, accumulation of dissolved molecular oxygen (O2), and an increase in pH (Gleitz et al. 1995; 1996; Günther et al. 1999; Papadimitriou et al. 2007). In closed bottle incubation experiments to study the growth physiology of ice algae under conditions simulating the sea ice interior during summer, Gleitz et al. (1996) demonstrated that photosynthetic carbon assimilation by ice algae can result in profound biogeochemical changes in brine solutions, which in turn have significant consequences for the sea ice biota. Other studies investigated the effects of temperature and salinity on ice algal growth both in laboratory experiments and in situ (e.g. Aletsee and Jahnke 1992; Arrigo and Sullivan 1992; Cota and Smith 1991; Mock 2002; Ralph et al. 2005, 2007), but simultaneous measurements of biogeochemical properties (e.g. inorganic nutrients and O2) of natural brines and photosynthetic parameters of the associated algal communities in sea ice are scarce.

Pulse amplitude modulation (PAM) fluorometry has been widely used to investigate the photophysiology of aquatic plants. The technique allows determination of the maximum quantum yield of photosystem II (PSII) for dark-adapted samples. Using rapid light curves (RLCs), comparable to classical photosynthesis versus irradiance (PE) curves, it is also possible to estimate the light-limited photosynthetic efficiency (α), maximum relative electron transport rate (rETRmax), and the photoacclimation index Ek of PSII. Recently, several studies have used PAM fluorometry to investigate the photophysiology of Antarctic bottom sea ice communities and their response to changing light levels, UV radiation and lowered salinities (McMinn et al. 2003; Ryan et al. 2004). Most studies have focused on the bottom ice community, while comparatively few investigated the community in the upper sea ice layers, which are subject to harsher environmental conditions and have recently been found to be well adapted to the cold temperatures and high salinities of their habitat in East Antarctic pack ice (Ralph et al. 2007). However, data on photosynthetic parameters combined with detailed physical and chemical information of the sea ice habitat are limited.

In this study, we present paired observations of brine biogeochemistry and ice algal photophysiology from surface and bottom habitats in natural sea ice. Our objective was to characterise the environmental conditions of these habitats during the winter-spring transition and investigate potential feedbacks between the photophysiology of ice algae and sea ice brine chemistry. Methodological problems that arise because physiological measurements provide a snapshot of ice algal activity whereas biogeochemical data provide time-integrated information on both biotic and abiotic processes are discussed.

Materials and methods

Site and sampling

Sea ice sampling was carried out in September and October 2006 in the north-western Weddell Sea during the Winter Weddell Outflow Study (WWOS) onboard the icebreaker RV Polarstern (expedition ANT XXIII-7). Fourteen ice floes were sampled for brine using the sackhole technique, in which brine is allowed to percolate into partial holes drilled into the ice surface. On each station, the snow was removed from a 1 m2 area, and using a 14 cm internal diameter Kovacs Mark V ice corer a sackhole (depth 20–60 cm) was drilled into the surface of the sea ice floe. Sufficient volume of brine accumulated in the lid-covered sackholes within 1 h. After determining the temperature of the brine in the sackholes, samples were collected in 60 ml borosilicate bottles and immediately fixed for O2 determination by Winkler titration. The remaining brine was transferred into an acid-washed polyethylene container and was transported to the onboard ship laboratories for determination of salinity and the concentration of inorganic nutrients, dissolved organic carbon (DOC), dissolved organic nitrogen (DON), and chlorophyll a (Chl a). A separate sample was collected for the determination of photosynthetic parameters (see below).

Bottom sea ice core sections were obtained from companion ice cores (9 cm diameter) on six of the sampled ice floes. Specifically, the lowermost 10 cm of two ice cores (named A and B) were cut with a stainless steel saw directly into acid-washed polyethylene jars, which were placed immediately into a dark insulated box. Bottom section A was used for determination of salinity and the concentration of inorganic nutrients, DOC, DON and Chl a, following melting at 4°C in the dark. Bottom section B was used for the determination of ice algal photosynthetic parameters. In order to get a liquid sample without melting the sea ice to avoid osmotic and temperature stress (Garrison and Buck 1986; Ryan et al. 2004), 80 ml of sterile-filtered (0.2 μm pore size) in situ temperature under-ice water were added to the crushed ice section and the algae were washed out of the ice matrix by carefully shaking the mixture for 2 min.

The ice temperature in the bottom ice core sections was measured immediately after coring from a small hole drilled in the ice core at 5 cm distance from the bottom surface. Sampling locations are given in Table 1, with station numbers representing the day of the year, i.e. 251 = 8 September 2006. Snow thickness was measured with a ruler at each site prior to sampling.
Table 1

Station number (day of the year), latitude, longitude, ice thickness, snow thickness, brine temperature in sackholes, salinity of collected brine (brine salinity), chlorophyll a concentration (Chl a) in collected brine, dissolved oxygen concentration in collected brine, ice temperature (measured 5 cm from bottom of floe), ice bulk salinity (of bottom 10 cm section), brine salinity* [calculated as function of ice temperature according to Assur (1958)], and Chl a* (chlorophyll a concentration in brine of bottom 10 cm section based on brine volume calculations from ice temperature and bulk salinity according to Frankenstein and Garner (1967) for the ice floes sampled during expedition ANT XXIII-7 (for details see text))

Station (day of the year)

Latitude south

Longitude west

Ice thickness (cm)

Snow thickness (cm)

Sackhole samples

Bottom ice samples

Brine temperature (°C)

Brine salinity

Chl a (μg l−1 brine)

Dissolved oxygen (μmol kg−1 brine)

Ice temperature (°C)

Ice bulk salinity

Brine salinity*

Chl a* (μg l−1 brine)































































































































































nd not determined, Empty cells not sampled

Physical and chemical parameters

Brine and bulk ice salinity, the latter from melted ice core bottom sections, were determined using a WTW Tetraconn 325 conductivity meter. The brine salinity in the bottom ice sections was calculated as function of ice temperature (Assur 1958), and the brine volume was calculated from bulk ice temperature and salinity according to Frankenstein and Garner (1967). The concentration of O2 in sackhole brines was measured by automated Winkler titration (Papadimitriou et al. 2007). The concentrations of nitrate plus nitrite [hereafter, nitrate (NO3)], dissolved inorganic phosphorus (DIP), silicic acid (Si(OH)4), dissolved ammonium (NH4+), DOC and DON were determined as described in Papadimitriou et al. (2007). For the determination of Chl a concentrations, subsamples of collected brines and melted ice core segments were filtered onto Whatman GF/F filters and kept frozen (−30°C) until analysis with a Turner Designs 10-AU fluorometer according to Arar and Collins (1997).

Photosynthetic parameters

Photosynthetic parameters of sea ice algae were determined from both brine and bottom ice samples using a Walz Water-PAM fluorometer (Walz GmbH, Effeltrich, Germany). The samples were protected from light during sampling and handling using non-transparent containers, a black tarp and working in darkened laboratories, and were dark adapted at in situ temperature for 20 min prior to fluorometric measurements. PAM fluorometers allow the determination of the maximum quantum yield (Fv/Fm) and relative electron transfer rate (rETR) of PSII in photosynthesis. RLC programs provide additional information on the photosynthetic efficiency of PSII and the photoadaptive state of the algae. A weak measuring light (0.15 μmol photons m−2 s−1) was used to measure an initial fluorescence response (F0, open PSII reaction centres), while a saturating pulse (>3,000 μmol photons m−2 s−1) was used to measure the maximum steady state fluorescence [(Fm) closed PSII reaction centres]. Red light emitting diodes provided light used in the RLCs at intensities of 0, 33, 46, 67, 101, 148, 230, 342 and 488 μmol photons m−2 s−1 for the bottom ice sections, and 0, 67, 101, 148, 230, 342, 488, 682 and 1,134 μmol photons m−2 s−1 for the sackhole brine samples. RLCs were generated from five subsamples, and the RLCs were individually analysed for the determination of photophysiological parameters. The relative electron transport rate (rETR) was calculated according to the formula rETR = Fv/Fm × PAR (Schreiber 2004). The RLCs were described using characteristic parameters, such as Ek, α and rETRmax (McMinn et al. 2003; Ralph and Gademann 2005). To determine these parameters, the RLC data were fitted to a double exponential decay function with a Marquardt–Levenberg regression algorithm (Platt et al. 1980) using the non-linear regression tool of SPSS 16.0 (2007) for Macintosh by SPSS Inc. RLCs were discarded from analysis when the Marquardt-Levenberg regression algorithm used for the estimation of parameters showed an asymptotic standard error >15%. In the absence of photoinhibition, the function can be simplified to a standard rectangular hyperbola with an asymptotic rETRmax value (Harrison and Platt 1986; McMinn et al. 2003):
$$ {\text{rETR}} = {\text{rETR}}_{\max } \left( {1 - {\text{e}}^{{ - \frac{\alpha E}{{{\text{rETR}}_{\max } }}}} } \right) $$
where rETR is the relative electron transport rate at a given irradiance, rETRmax is the rETR at saturating light, α is the initial slope of the RLC before the onset of saturation also called efficiency of rETR, and E is the irradiance (provided by a LED array with a peak emission at 650–660 nm). The photoacclimation index (Ek) was calculated according to the formula Ek = rETRmax/α (e.g. McMinn et al. 2003; Ralph and Gademann 2005).

Non-parametric Spearman’s rank tests were used to explore correlations between variables. The non-parametric Mann–Whitney U-test was used to determine significant differences between median values of non-normal distributed data. All statistical tests were computed using SPSS 16.0 (2007) for Macintosh by SPSS Inc.


Physical parameters

Ice and snow thickness varied between 63 and 218 cm (average 128 ± 47 cm), and 3 and 91 cm (average 28 ± 25 cm), respectively (Table 1). Most of the sampled sea ice was first-year sea ice with the exception of stations 262 and 271, which were multi-year sea ice characterised by bulk ice salinities below 1 in the upper parts of the ice column (data not shown). The temperature of the sackhole brines ranged from −7.4 to −3.8°C, while the temperature of the bottom ice core sections was warmer, ranging from −2.1 to −1.9°C and resembling the stable temperature conditions at the ice–water interface. Similarly to the difference in temperature between the sackhole brine and the bottom ice layer, the sackhole brine salinity ranged from 61.4 to 118.0, and was higher than the calculated salinity of the brine in the bottom ice core sections, which ranged from 33.9 to 37.4 (Table 1).

Biogeochemical parameters

The concentrations of inorganic nutrients, DOC and DON in sea ice brines are shown in Fig. 1. Solutes behave conservatively during the freezing and melting of seawater when these physical processes are the only ones controlling their distribution within sea ice. The dashed lines in Fig. 1 denote conservative behaviour during the freezing and the melting of sea ice derived from winter surface water [NO3 = 30 ± 1 μmol kg−1, DIP = 2.1 ± 0.1 μmol kg−1, Si(OH)4 = 69 ± 6 μmol kg−1 (measured during this study) and NH4+ = 0.3 ± 0.2 μmol kg−1, DOC = 51 ± 15 μmol kg−1 and DON = 3.6 ± 3.2 μmol kg−1 (measured during the ISPOL study, Papadimitriou et al. 2007)]. The NO3 concentration in the brines ranged from 0.5 to 76.3 μmol kg−1, with all sackhole samples and some of the bottom ice samples lying below the conservative line, indicating depletion relative to the composition of surface seawater (Fig. 1a). The Si(OH)4 concentrations (range 15–285 μmol kg−1) were depleted in all the bottom ice samples and most of the sackhole samples, while the DIP concentrations (range 0.2–7.0 μmol kg−1) were depleted in the sackhole brines but were at or above the conservative line in the bottom ice layer (Fig. 1b, c). The NH4+ concentrations exhibited a wide range (0.3–23.0 μmol kg−1) and were elevated in both the sackhole brines and the bottom ice samples relative to surface seawater. The DOC and DON concentrations ranged from 45 to 707 μmol kg−1 and 6 to 67 μmol kg−1, respectively (Fig. 1e, f). The DOC concentrations were elevated in both the sackhole brines and the bottom ice samples, indicating substantial enrichment in both types of sample relative to surface seawater. In comparison with conservative behaviour, DON was enriched in all the bottom ice samples and in about half of the sackhole brine samples. DOC/DON ratios ranged from 10.5 to 35.4 (average 18.2 ± 6.5) and 3.3 to 10.4 (average 6.9 ± 2.3) in the sackhole and bottom samples, respectively.
Fig. 1

Concentration of inorganic nutrients, dissolved organic carbon (DOC) and dissolved organic nitrogen (DON) in sea ice brines versus salinity of brine. Open circles denote bottom ice samples for which brine concentrations were calculated from brine volume estimates (Frankenstein and Garner 1967, for details see text). Filled circles denote brines samples collected from sackholes. Dotted lines indicate nutrient concentrations to be expected from surface water values (average ± SD) if that nutrient was conservative with salinity. Note logarithmic scale in d

The concentration of O2 in the sackhole brines ranged from 159 to 413 μmol kg−1 (Table 1). The percent oxygen saturation relative to equilibrium with air in sackhole brines is presented in Fig. 2. In about half of the sampled ice floes O2 in sea ice brines showed a moderate supersaturation of 129 ± 18% (n = 8), while the remaining ice floes showed O2 concentrations at or below air saturation with an average of 90 ± 15% (n = 6).
Fig. 2

Oxygen saturation relative to equilibrium with air in sea ice brines collected from sackholes. The dashed line indicates 100% oxygen saturation

The concentrations of total inorganic nitrogen (TIN = NO3 ± NH4+), DIP and SI(OH)4 were normalised to a salinity of 34 as in Gleitz et al. (1995), with pairwise distributions relative to TIN shown in Fig. 3. The resulting ratios, TIN/DIP and TIN/Si(OH)4, ranged from 1.6 to 30.3 (average 14.7 ± 9.7) and 0.2 to 7.0 (average 1.8 ± 2.0), respectively.
Fig. 3

Salinity (S = 34) normalised a dissolved inorganic phosphate (DIP), and b silicic acid Si(OH)4 concentration versus salinity normalised total inorganic nitrogen (TIN) concentration. Dotted lines in (a) indicates N:P ratio of 16:1 and in (b) indicates N:Si ratio of 16:18.5 for sea ice diatoms (Günther et al. 1999)

The concentration of Chl a was generally low in the sackhole brines, ranging from 0.3 to 11.8 μg l−1. The Chl a concentration in the brines of the bottom ice sections showed a much higher variation, exceeded the concentrations in the sackhole brines at all except one station, and varied between 4.9 and 435.5 μg l−1 (Table 1).

Photophysiological parameters

Fv/Fm of the sackhole samples showed a wide range with a minimum of 0.101 and a maximum of 0.500 (average 0.284 ± 0.132, median 0.331) (Table 2). The Fv/Fm values of bottom ice algae ranged between 0.235 and 0.595 (average 0.368 ± 0.127, median 0.350), and there was no significant difference in the median Fv/Fm values of the two habitats (Mann–Whitney U-test P = 0.284). The Fv/Fm tended to increase with temperature and to decrease with brine salinity, but the relationships were not statistically significant (Spearman’s rank correlations: temperature ρ = 0.391, P = 0.089; brine salinity ρ = −4.05, P = 0.085). Additionally, no significant correlations were found between Fv/Fm and brine oxygen concentration (Spearman’s rank correlation ρ = −0.354, P = 0.215). The maximum relative electron transport rates (rETRmax) in the sackhole brines and bottom ice samples ranged from 3.74 to 20.58 (average 9.48 ± 4.92, median 7.51) and 3.03 to 9.25 (average 5.33 ± 2.26, median 4.71), respectively. The median rETRmax values of the two habitats were significantly different (Mann–Whitney U-test P = 0.039). Relative ETRmax values were not correlated with temperature, brine salinity and oxygen concentration (Spearman’s rank correlations: temperature ρ = −0.507, P = 0.023; brine salinity ρ = 0.407, P = 0.075; oxygen concentration ρ = 0.292, P = 0.311). The light-limited rETR efficiency (α) of ice algae collected from the sackholes ranged from 0.016 to 0.142 (average 0.069 ± 0.031, median 0.083), while bottom ice algae had α values in the range of 0.038 to 0.130 (average 0.085 ± 0.032, median 0.080). The median values of α showed no significant difference between the habitats (Mann–Whitney U-test P = 0.458) and α was not correlated to temperature, brine salinity and oxygen concentration (Spearman’s rank correlations: temperature ρ = 0.161, P = 0.497; salinity ρ = −2.43, P = 0.303; oxygen ρ = −1.74, P = 0.553). The photoacclimation indices (Ek) of sea ice algae were statistically significant higher in the sackhole samples (range 55.36–313.27 μmol photons m−1 s−1, average 162.09 ± 76.40 μmol photons m−1 s−1, median 164.25 μmol photons m−1 s−1) from the upper parts of the sea ice when compared to the ice algal communities at the bottom of the sea ice (range 40.90–115.26 μmol photons m−1 s−1, average 66.75 ± 26.08 μmol photons m−1 s−1, median 60.39 μmol photons m−1 s−1, Mann–Whitney U-test P = 0.006).
Table 2

Station number (day of the year), photosynthetic yield of photosystem II (Fv/Fm), maximum relative electron transport rate (ETRmax), the initial slope of the rapid light curve (α) and the photoacclimation index (Ek) for sackhole and bottom ice samples taken during ANT XXIII-7





Ek (μmol photons m−2 s−1)

Sackhole samples


0.500 ± 0.034

11.34 ± 2.02

0.142 ± 0.012

81.09 ± 20.22


0.362 ± 0.034

5.53 ± 0.30

0.084 ± 0.017

67.13 ± 12.25


0.427 ± 0.011

10.54 ± 2.36

0.082 ± 0.012

131.71 ± 36.03


0.140 ± 0.025

8.17 ± 0.68

0.047 ± 0.008

180.06 ± 37.53


0.136 ± 0.030

5.89 ± 0.98

0.036 ± 0.003

166.96 ± 41.14


0.354 ± 0.006

3.74 ± 0.17

0.016 ± 0.001

229.39 ± 18.25


0.385 ± 0.012

5.61 ± 0.33

0.102 ± 0.004

55.36 ± 5.18


0.341 ± 0.005

20.58 ± 1.13

0.092 ± 0.007

226.10 ± 23.83


0.143 ± 0.009

5.95 ± 0.64

0.026 ± 0.006

243.15 ± 81.08


0.321 ± 0.023

5.56 ± 0.70

0.083 ± 0.005

67.07 ± 9.50


0.367 ± 0.129

14.77 ± 1.92

0.084 ± 0.020

185.69 ± 61.59


0.113 ± 0.035

12.80 ± 2.63

0.041 ± 0.004

313.27 ± 85.52


0.101 ± 0.026

6.84 ± 0.50

0.043 ± 0.005

160.71 ± 22.10


0.279 ± 0.058

15.37 ± 1.61

0.096 ± 0.010

161.53 ± 24.61

Bottom ice samples


0.282 ± 0.010

3.03 ± 0.21

0.075 ± 0.006

40.90 ± 6.00


0.595 ± 0.053

9.25 ± 0.47

0.130 ± 0.027

72.89 ± 13.37


0.389 ± 0.032

3.75 ± 0.54

0.074 ± 0.007

50.64 ± 4.15


0.311 ± 0.052

4.38 ± 0.23

0.038 ± 0.003

115.26 ± 4.13


0.235 ± 0.015

5.05 ± 0.73

0.085 ± 0.006

59.85 ± 11.95


0.396 ± 0.018

6.52 ± 0.80

0.111 ± 0.019

60.93 ± 19.50

Data represent average ± SD values


Our study provides information on paired observations of brine biogeochemistry and ice algal photophysiology from surface and bottom habitats in Weddell Sea ice during early spring. The study adds to the growing body of literature on PAM measurements of ice-associated algae from various Antarctic sea ice habitats and provides the first PAM data for Weddell Sea pack ice interior and bottom algal communities.

Sampled ice floes comprised both first-year sea ice and second-year sea ice, with average ice and snow thicknesses similar to the averages reported in the ASPeCt data base for springtime western Weddell Sea ice [ASPeCt data: ice thickness 1.33 ± 0.85 cm and snow thickness 0.24 ± 0.19 cm (Worby et al. 2008)], indicating that we sampled sea ice that was characteristic for the study area. Low brine temperatures and corresponding high brine salinities indicated the low surfaces temperatures of the sampled sea ice. Most sackhole brine temperatures were below −5°C suggesting that the upper parts of the sea ice had brine volume fractions below the theoretical percolation threshold of sea ice (Golden et al. 1998), and can be considered as a semi-enclosed system in which slow diffusive fluxes and in situ heterotrophic remineralisation are the only mechanisms to resupply inorganic nutrients to autotrophic organisms. In contrast, the bottom sections showed relatively warm temperatures and high bulk salinities, favouring the exchange of dissolved constituents between the sea ice and the underlying seawater. Replenishment of nutrients from the under-ice water is the likely reason for the high Chl a concentrations observed in the sea ice bottom sections.

As a result of the cold ice temperatures reducing brine volumes and concentrating dissolved constituents, nutrient concentrations measured in the brines in the present study were high when compared to those reported previously for brine from various ice types in the Weddell Sea during summer, autumn, winter and late spring (Gleitz and Thomas 1993; Gleitz et al. 1995; Papadimitriou et al. 2007). Inorganic nutrient concentrations behave conservatively during the freezing and melting of sea water in the absence of biological activity, such that their concentrations should be a linear function of salinity. In general, the inorganic nutrient concentrations in the sackhole brines indicated depletion in NO3, DIP and partly Si(OH)4, while the brine in the bottom ice sections was depleted in Si(OH)4. The reduced nutrient concentrations indicate that autotrophic nutrient consumption had occurred in both habitats. The NH4+ concentrations indicated enrichment relative to surface seawater, suggesting grazing activity and heterotrophic nitrogen remineralisation in both the sea ice interior and bottom horizons. Kristiansen et al. (1998) proposed that early season Antarctic infiltration algal communities utilise primarily nitrate, with the rate of NH4+ uptake increasing only later in the growing season. The high NH4+ concentrations measured in the present study provided a readily available nitrogen pool for the sea ice autotrophs and are assumed to have been used by the algal communities in both the upper and lower parts of the sea ice. While NO3 was depleted in comparison to DIP and Si(OH)4 in both the interior and the bottom of the sea ice (Fig. 3), concentrations were typically above the half-saturation constants reported for Antarctic phytoplankton (≤4 μM; Reay et al. 1999) and thus not likely to limit algal growth during the time of sampling. In summary, the inorganic nutrient concentrations suggest that autotrophic nutrient drawdown as well as heterotrophic remineralisation had occurred both in the interior and at the bottom of the sampled sea ice floes.

Accumulation of dissolved organic matter (DOM) as DOC and DON was observed in most samples (Fig. 1). The DOC/DON ratios in the sackhole brines (average 18.2) exceeded those measured in the bottom ice samples at all stations, demonstrating more carbon-enriched DOM in the upper sea ice layers. High DOC/DON ratios in sea ice have been explained by the uncoupling of the carbon and nitrogen metabolisms, as well as by production of carbon-rich extracellular polymeric substances by ice algae (Krembs et al. 2002; Meiners et al. 2004; Thomas and Papadimitriou 2003). The bottom sections showed DOM with low C/N ratios (average 6.9), close to the Redfield value of 6.6, indicating that this material was more biologically available than both the DOM in the upper parts of the sea ice and the DOM in the winter surface water with an average DOC/DON ratio of 14.2.

Large O2 supersaturation (up to 200%) of sea ice brines due to high photosynthetic activity has been reported previously (Delille et al. 2007; Gleitz et al. 1995; Papadimitriou et al. 2007). In the present study sea ice brine oxygen showed moderate supersaturation in about half of the sea ice brines, and saturations at or below 100% air saturation in the remaining samples. Hyperoxia impacts negatively on ice algal physiology due to both competition of carboxylase and oxygenase reactions at the site of RUBISCO and damaging effects of reactive oxygen species on chloroplasts and other cellular components (McMinn et al. 2005; Raven et al. 1994; Thomas and Dieckmann 2002). In the present study no statistically significant correlations between oxygen concentration and the photophysiological parameters Fv/Fm, rETRmax and α were found, and no clear conclusion on the impacts of oxygen concentration on ice algal physiology in natural sea ice can be drawn from our limited data set. Modifications of brine oxygen concentrations by abiotic processes, i.e. degassing (Papadimitriou et al. 2007), as well as the generally moderate oxygen saturation levels are likely to have contributed to masking the relationships between brine oxygen concentrations and photosynthetic parameters.

Most studies of the primary production and physiology of Antarctic sea ice algae have been based on 14C-tracer methods or oxygen exchange techniques using closed chambers or microsensors. Only recently PAM and RLC fluorometric methods have been applied to quantify photosynthetic parameters of Antarctic sea ice algae. Since ice algal physiology changes throughout the daily cycle of irradiance, RLCs and the derived parameters α, rETRmax and Ek reflect a snapshot of the photophysiological condition of the algae at a distinct time in the diurnal cycle. In the present study the maximum quantum yield (Fv/Fm) of Chl a fluorescence obtained from dark-adapted samples averaged 0.284 ± 0.132 in the sackhole samples and 0.368 ± 0.127 in the bottom sections. These values are lower than the theoretical optimum of 0.65 for phytoplankton (Schreiber 2004) and also at the lower end of the range of Fv/Fm values reported from different ice algal studies (Arrigo et al. 2003; McMinn and Hegseth 2004; McMinn et al. 2008). Fv/Fm values in the present study are also low when compared to McMinn et al. (2003, 2007) reporting average Fv/Fm values of 0.45 ± 0.15 and 0.470 ± 0.041 for springtime bottom ice communities from Antarctic coastal fast ice (McMurdo Sound) and East Antarctic offshore pack ice, respectively.

Temperature and brine salinity are coupled in sea ice (Assur 1958). Complex patterns of inhibition of photosynthesis by cold temperatures and hypersaline conditions have been reported for Antarctic ice algae (e.g. Arrigo and Sullivan 1992; Kirst and Wiencke 1995; Ralph et al. 2007; Ryan et al. 2004). In general, the studies report on impaired photosynthetic activity of ice algae at salinities >60 and temperatures below −5 to −10°C due to cold shock inhibition (e.g. Ralph et al. 2005). While our limited data set did not show significant correlations between photosynthetic parameters and temperature, and salinity, these previous studies suggest that cold temperatures and high brine salinities found in the sackhole brines in the present study are a likely factor for the reduced Fv/Fm values measured in these samples, while bottom ice communities with brine temperatures above −2.2°C and brine salinities below 40 were most probably unaffected by temperature and salinity stress.

The rETR efficiency under light limitation (α) was low both for the internal and bottom ice communities, and there was no statistically significant difference between the habitats. While communities in the ice interior were most probably affected by cold temperatures, high salinities, and elevated oxygen concentrations, the bottom communities are assumed to have been light-limited. Downwelling irradiance was not measured in the present study but we estimated irradiances at 40 cm ice depth and at the bottom of the ice floes according to Beer–Lambert equations (e.g. Gradinger 2008) for clear sky conditions assuming a surface irradiance of 1,000 μmol photons m−2 s−1, a surface albedo of 0.8 and attenuation coefficients for snow and ice of 10 and 1.5 m−1, respectively (Perovich 1996 and references therein). These upper estimates for the irradiance levels were <0.1–81 photons m−2 s−1 (average 28 photons m−2 s−1, n = 14) and 0.4–28 photons m−2 s−1 (average 9 photons m−2 s−1, n = 6) for the surface (40 cm depth) and bottom layers, respectively. Average Ek values for sackhole (Ek = 162 photons m−2 s−1) and bottom (Ek = 67 photons m−2 s−1) samples were much higher than these estimates and indicate low levels of light acclimation in both habitats. This finding is in agreement with observations by Gradinger (2008) reporting on Ek values that are substantially above in situ irradiance levels for Arctic sea ice algae. Nevertheless, the average Ek value of 67 μmol photons m−2 s−1 found for the bottom communities in our study is only slightly elevated when compared to Ek values of 19–44 μmol photons m−2 s−1 reported for East Antarctic pack ice bottom communities (McMinn et al. 2007). However McMinn et al. (2003) report on the diurnal variation of Ek and rETR of Antarctic fast ice algal communities. Our samples were taken at different times during the day and thus direct comparison to other studies is difficult when samples are not taken at a standardised time (McMinn et al. 2003). While our PAM-RLC data set is relatively small, it nevertheless provides the first estimates for the range for α, Ek and rETRmax for Weddell Sea ice communities during early spring. The observed sea ice algal communities showed moderate levels of photophysiological parameters and limited light acclimation both at the sea ice bottom as well as in the cold, high saline and partly oxygen supersaturated sea ice brines in the upper parts of the ice floes.

The observed lack of correlations between the biogeochemical parameters and ice algal photosynthetic parameters was most probably the result of temporally decoupled physical and biological processes, and was also influenced by measurement techniques. While PAM fluorometric measurements from sackhole brines provide a snapshot of algal photosynthetic parameters (Glud et al. 2002), measurements of dissolved constituents provide time-integrated information and combine effects from a suite of physical (e.g. gas exchange and brine drainage) and biogeochemical processes (e.g. photosynthesis and respiration). In addition, while the sackhole technique provides quantitative sampling of dissolved constituents (Papadimitriou et al. 2007) the method underestimates the particulate fraction which partly remains trapped in the brine channel/ice matrix. This is also true for our method to extract bottom ice algae focusing on reduction of temperature and osmotic stress. Weissenberger (1992) reports that 70–90% of the Chl a remains trapped in ice cores even after centrifugation. Assuming that entrapment of algae in brine channels is species specific, i.e. due to differences in cell size, cell surface properties and surface association, also our PAM measurements may not be absolutely representative for the ice algal communities under investigation. Furthermore ice algae tend to easily form aggregates (Riebesell et al. 1991) and aggregates in the PAM samples may have contributed to variability in the replicate photophysiological measurements at individual stations (Table 2). This may have additionally contributed to the masking of relationships between biogeochemical and photophysiological parameters. Microscopy PAM measurements (e.g. McMinn et al. 2008) of individual cells from the same habitat may be a way forward to understand within-sample-variability of photosynthetic measurements.

In conclusion our study provides a characterisation of the biogeochemical conditions in Weddell Sea ice during early spring indicating autotrophic nutrient drawdown as well as heterotrophic remineralisation in both sea ice bottom and interior habitats. PAM fluorometric measurements showed moderate activity of ice algae both in the interior and at the bottom of the sea ice floes. The present study highlights problems in matching biogeochemical and photophysiological measurements in natural sea ice. Larger data sets as well as integrated time series pairing snap shot physiological measurements and time-integrated biogeochemical measurements are needed to further our understanding of coupled sea ice biogeochemical and ice algal photophysiological processes in natural sea ice.


We thank the captain and crew of RV Polarstern for their cooperation during cruise ANT XXIII-7, as well as those who supported our work in the field. Our laboratory work on the ship was supported by E. Allhusen, C. Uhlig, M. Kramer and R. Kiko. We thank M. Raateoja, K. Ryan and an anonymous reviewer for their helpful comments on an earlier draft of the manuscript. The work of DNT, LN and SP was funded by NERC, UK. This work was supported by the Australian Government’s Cooperative Research Centre program through the Antarctic Climate and Ecosystems Cooperative Research Centre (ACE CRC).

Copyright information

© Springer-Verlag 2009